ANCESTRAL INFLUENCE 



525 



is analogous to sexual inheritance, as Dr Reid has argued. Mendelian 

 inheritance deals with the combination of two characters, one of which 

 is patent and the other latent or non-existent. It could be represented 

 graphically by the same artifice used for sex. But properly to develop 

 the consequences requires a more careful study of the phenomena than 

 the present writer has been able to undertake. 



27. Ancestral resemblance. Let us now return to those qualities 

 which are transmitted non-sexually. We have hitherto dealt simply with 

 the resemblance of children to their 

 immediate parents ; but we know that 

 inheritance is not from parents alone. 

 From wide experience we know that 

 a child has a tendency to resemble, 

 not only its immediate parents, but its 

 ancestors, even when the parents differ 

 from them (IX.). 



Hence if OP be the ancestral type, 



FIG. 10. 



and OQ the path of a parent, the 

 path of the child, Oq, will tend to lie 

 between OQ and OP : but there is 

 no reason why it should be midway between OQ and OP : probably it 

 lies nearer OQ, though we may not be able to deduce the numerical law 

 from theoretical considerations. Sir Francis Gal ton found in 1889 that 

 in certain cases Pq= 2 Qq : or Pq = f PQ : i.e. the deviation of a child 

 was f of the deviation of a parent: and this may quite possibly be a 

 general law deducible from theoretical considerations. On the ocher 

 hand, the numerical factor may differ in every case. For the present, we 

 shall merely use Sir Francis Galton's result for the purpose of illustration 

 without attempting to explain it, or inquire into its generality. 



28. Corrected statement of sexual 

 effect. Thus if one parent alone (OQ) 

 were concerned, the child would tend to 

 follow the path Oq, where Pq = f PQ. 

 If the other parent (OR) were alone 

 concerned, the child's path would tend 

 to be Or, where Pr=f PR. When 

 both parents are concerned, the child's 

 path will therefore tend to lie midway 

 between Oq and Or, not midway be- 

 - tween OQ and OR as previously stated. 

 The deviation from OP will consequently 

 be of that previously stated, which 

 we found was, on the average, one half that of the parents. Hence the 

 deviation will be reduced every generation, not in the ratio \ which is 

 the sexual effect, but in the ratio f X J = J, which is the combined sexual 

 and ancestral effect. Returning to the analogy of the snail, if at the end of 

 each hour one person sets him back to half his distance, and another 

 person to f of that half, he will always be set back to J of his total 



FIG. ii. 



