A. Barrington, a. Lee and K. Pearson 
255 
The mean of 12 fourfold tables, three for each parental table, by methods 
of division A, B, C above, was -526. We think we may therefore safely say that 
the 36-fold contingency table leads to results in quite good agreement with the 
mean of the older process of dealing with quantities not quantitatively measurable. 
Nor are the individual results which by the fourfold method are : Sire and 
Dog, -538 ; Sire and Bitch, -615 ; Dam and Dog, -467 ; Dam and Bitch, -482, widely 
different. Nor do we find any sensible difference if we consider the inheritance 
of ' red pigment ' and of ' black pigment,' the mean parental coefficients being '520 
and "529 for the two cases respectively. Lastly, the mean parental correlation 
from four 25-fold and four 16-fold contingency tables was "492, somewhat less 
than the value as given by the fourfold method or the 36-fold contingency method, 
which, as we have seen, are in good agreement. We believe that "-53 represents 
extremely well the average parental relation in pigment in the case of greyhounds. 
We compare this in the following table with previous pigmentation i-esults : 
TABLE V. 
Direct Inheritance of P if/ mentation. 
Man 
Eye-Colour 
Horse 
Coat-Colour 
Dog, 
Basset Hound 
Coat-Colour 
Dog, 
Grej'hound 
Coat-Colour 
Mean 
Father and Son 
■5.50 
•491 
?* 
•512 
•518 
Father and Daughter 
•437 
•542 
1 
•579 
•519 
Mother and Son 
•482 
•486 
•508 
•505 
•495 
Mother and Daughter 
•510 
•567 
•.544 
•532 
•538 
Mean 
•495 
•522 
•526 
•5.32 
•518 
This table seems to indicate that there is no evidence at present to show 
that pigmentation inheritance differs in intensity in man, horse or dog, nor is 
there evidence to indicate any persistent preponderance of one sex. We should 
notice, however, that in all four cases the daughter is more like the mother than 
the son is. On the other hand the daughter is more like the father than the 
son is in horse and dog, the sex similarity being only preserved in the case of 
eye-colour in man. We may safely say that the inheritance of pigmentation 
cannot differ widely from 50 for all our series. 
If we turn to parental inheritance in the case of dogs selected for record we 
see that the correlation has been reduced, a result which is the general rule 
whenever selection takes place. The reducing factor is "875. Being completely 
ignorant of the relative standard deviations of the material under consideration, 
we can hardly reach a more scientific measure of the selection that takes place in 
record than this reducing factor. But its value is sufficient to show us that our 
original hesitation to use the greyhound stud-book was justifiable ; the correlation 
* See R. S. Proc. Vol. 66, p. 159 ftn. Increased information as to inheritance in dogs contirms the 
view that the sires given in the Basset Stud-book are unreliable. 
