A. Barrington and K. Pearson 
451 
more potent than two changes (and than either one or three). There does not, 
however, seem to be much marked difference between one and three changes. On 
the whole it seems probable that a change of sex does produce some slight lessening 
in the ancestral resemblance beyond the parents; it is marked in eye-colour in 
man, and just appears in the coat-colours of greyhound and shorthorn, although 
if it exists in coat-colour of horses it has been screened by the probable errors of 
the results. 
The mean result for eye-colour in man for great-grandparental resemblance is 
•19, and for horse's coat-colour '20, both results sensibly higher again than that for 
shorthorns. 
Still one further ancestral resemblance has been worked out, namely, that 
between great-great-grandparents and the offspring. But the complete set of 
32 tables has been replaced by four dealing only with each sex in the ancestry 
and the offspring regardless of the exact line of descent. The results are given 
below* : 
TABLE VII. Great-Great-Grandparent and Offspring. 
Great-Great-Grand.sire and $ Offspring 
•09 
Great-Great-Grandsire and $ Offspring 
■14 
Great-Great-Grandam and j Offspring 
•13 
Great-Great-Grandam and ^ Offspring 
•13 
Mean =-12 
The only comparable result is the combined table for all sexes provided by 
Dr Alice Leef in the case of coat-colour in horses, and this gave for the great- 
great-grandparent and offspring the value "IS, thus agreeing with the previous 
cases in giving a higher value than the shorthorn results do. 
We can now put together our results for ancestry in shorthorns and place 
alongside them those for other pigmentation researches : 
TABLE VIII. Ancestral Resemblance. 
Species 
Man + 
HorseJ 
Basset § 
Hound 
Greyhound || 
Shorthorn 
Mean 
Result 
Character 
Eye- 
Colour 
Coat- 
Colour 
Coat- 
Colour 
Coat-Colour 
Coat-Colour 
Pigmen- 
tation 
Method employed 
Fourfold 
Fourfold 
Fourfold 
36-fold 
25-fold 
Various 
Table 
Table 
Table 
Contingency 
Contingency 
Parent 
•49 
•52 
•53 
.53 
•40 
•49 
Grandparent 
■32 
•30 
•22 
•33 
•20 
•27 
Great-Grandparent 
■19 
•19 
•17 
•18 
Great-Great-Grandparent 
•15 
•12 
•13 
* Given as Tables X, Y, Z, fi, in the Appendix, f Biometrika, Vol. ii. p. 235. t Ibid. Vol. ii. p. 222. 
§ R. S. Proc, Vol. Lxvi. p. 157, omitting the cases of inheritance through the sire, as there can be 
little doubt that the sires are unreliable. |] Biometrika, Vol iii. pp. 254-8. 
57—2 
